KOL Corner: Can Complement Inhibitors Treat Macular Degeneration?

[creativ_pullleft colour=”light-gray” colour_custom=”” text=”Episode 080″]

The complement system is a promising target for drugs being developed to treat macular degeneration. But success has been elusive. A leading KOL and two CEOs explain what the future of the field may be.

Podcast Transcript

Tom Salemi: Hi, everyone, welcome back to the OIS Podcast. This is your host, Tom Salemi. Very happy to have you here. We’re going to take a seat in the KOL Corner today. Now this time we’re looking at the potential therapies that are targeting the complement cascade as a treatment for macular degeneration. This discussion sort of centers around some disappointing results that were released at Angiogenesis back in February, specifically Novartis’ LFG 316 in a phase 2b study was shown that it didn’t perform any better than the sham therapy in targeting geography atrophy, lesion size. Novartis still has plans for LFG 316, but the news led us to look at a few other approaches that target different elements of the complement cascade. But first we’re going to hear from a KOL in the space, Dr. Jay Duker, the director of the New England Eye Center up here in Boston, gives us his overview of the various attempts to use the complement cascade as a way to combat macular degeneration.

TS: You were at Angiogenesis recently, and there were some interesting results there. I wonder do you have any principle takeaways from that event?

Jay Duker: There’s still quite a bit of science and industry interest in ophthalmology, specifically in retina, and not only in established pathways like the VEGF pathway, but establishing that new pathways be attacking those alternative pathways, we may actually improve disease as well. So that I think from the perspective of opportunities in the retina space, it’s still quite strong. In saying that, I don’t think that there was any real new, earth shattering information that was delivered at this Angiogenesis meeting. I think that we are going to have some later in the year. Ophthotech with their phase 3 trial hopefully will read out in the next year. I already mentioned some gene therapy trials that may read out later in this year. So that there’s some events that hold a lot of potential that are really upcoming. So we really look forward to hearing those results.

TS: And what was your takeaway from the Novartis results announcement?

JD: So specifically I think the Novartis result you’re talking about is they did a phase 2 trial for dry AMD patients with geographic atrophy. They have an antibody that supposedly blocks C5. Their name of it was LFG316. And what C5 does in the complement pathways, it’s cleaved by an enzyme into C5A and C5B, and those two cleavage molecules or products activate different parts of the complement cascade. So it was hoped that by blocking C5 in the eye, you could stop the progression of geographic atrophy. And it was a well-designed phase 2 trial, and unfortunately for the millions of patients who have dry AMD around the world, it didn’t appear that by directly blocking C5 cleavage with their antibody there was really any benefit. I think the results, again, are not particularly surprising. And I think there’s a few reasons why it might not have worked. I think – first of all, let me say I think that there’s overwhelming evidence that complement plays a role in macular degeneration, probably in both dry and wet macular degeneration. I think again the genetics, the basic science, the findings of membrane intact complex within drusen, we know that the complement system is a player. But the reason it didn’t show benefit, I think possibly that complement may be responsible for the initiation of macular degeneration, but not its perpetuation. So that if you’re going to block complement, you may have to do it early. I think another possibility is the study just didn’t go long enough. That if cells are programmed to die, the penumbra of cells at the edge of the geographic atrophy, they may – even if you block complement, those cells may be dying over the next 6 months to a year, and therefore you may just have to take the study out longer to show a benefit. On the other hand, I think that the real reason this didn’t work is just has to do with the stoichiometry of the target. C5, if you look at the complement cascade on a piece of paper, C5 looks like a great place to block it. But what you have to know is there is a lot of C5 in the body. For example, if you’re talking about a molecule like VEGF, when we measure VEGF in the eye, it’s pico milligram per ML. that’s the level of VEGF. With complement factors, you’re talking about micro milligram per ML. So it’s 100,000 to a million times more concentrated that VEGF for example. So when you think about that is how are you going to block all that C5? There’s just a lot of it. And so the other thing is because there’s no really good dry macular degeneration model to use, you can’t really test it in animals, so you have to test it in some other types of situations where your antibody may show an effect at blocking the lysis of red blood cells, for example, but really may not show that same effect in the eye. But I think this is just a stoichiometric problem. I think C5 on paper is a great target, but you’re just going to have to get a lot of anti-C5 drug in there constantly to damp down the complement cascade.

TS: And this has been a tough space for a number of companies. Are there other trials out there in C5 that you’re watching?

JD: Oh, yeah. I mean it’s tough. Complement has been an area of study for many, many years by many companies. And there are many systemic diseases including organ rejection and lupus and some important CNS diseases that are linked to the complement system. But if there’s so much complement in the body and there’s so many redundancies in the pathways, it’s a difficult problem. There’s not many sites of the complement pathway that I would say are easily drugable. But yeah, there are other people still trying to accomplish this. And I know Ophthotech’s got a phase 2 trial on Zimura, which is an aptamer against C5 that they are recruiting for. It’s interesting. Ophthotech’s approach has also been to test Zimura as a combination therapy with anti-VEGF in wet AMD. So not just as a geographic atrophy blocker, but as a possible help for wet AMD. And there’s some pretty good evidence that complement plays a role in corneal vascularization, at least it’s in initiation as well. So they had some interesting findings in their phase 1 study of the combo, and that may be how their C5 gets developed.

TS: And does that have anything to do with – I mean Novartis made a statement that they’re also going to introduce a combination therapy of their own with their LFG316 and properdin

JD: Right, which I believe is an anti-properdnia And so that’s an interesting approach, too. Again, I stated already the complement system is really complicated. And really to understand it, it’s really a three step system. There’s initiation of the complement system, what sets it off. And we know pretty well what sets it off, and things like oxidative stress, for example, is a big initiator of complement. Smoking, obesity, the inflammasome. All of those have been shown to initiate complement. The next part about it, though, is the amplification phase. And that’s what happens in the alternative pathway, where it’s like a feedback loop, where it revs up higher and higher. And those people who lack complement factor H have a more revved up alternative pathway. So properdin is something that allows the C3 enzyme that cleaves C3 to last longer. So it keeps the amplification loop going. And by blocking you may be able to damp down the amplification loop. But the idea here is because complement is so redundant, the complement system, there’s so many proteins, by hitting it in several places, you may be able to have a beneficial effect. So I look forward to the results. I certainly hope it works because we really need something for our dry AMD patients.

TS: The companies like Apellis and their C3 approach, I mean are they facing the same challenges? Or is that an entirely different –

JD: You know, again, they may be. They may be very well facing a stoichiometric problem, although there is some evidence I know from the Apellis drug that they developed originally, the POT-4 that there was some long term kind of slow release aspects to it. And so they may be able to overcome the stoichiometric problems by having a slow release type of version of the anti-C3. But once again, C3 is a very, very prevalent molecule in the body.

TS: As Jay Duker indicated, Ophthotech is also building a case for its own C5 approach, called Zimura. So we talked about Dr. Samir Patel. He of course is president and cofounder of Ophthotech, and was cofounder of Eyetech Pharmaceuticals before that. We talked to Dr. Patel about how that company views the Novartis news, although he didn’t get too in depth about talking about a competitor’s clinical news. More importantly, though, we focused on how Ophthotech views the complement cascade, and what its own plans are for Zimura. Let’s have a listen.

TS: Can you talk a bit about the complement cascade and how it fits into AMD? And where does Zimura sort of fit into stopping that progression?

Samir Patel: Sure so really, generally speaking, there are multiple ways to look at the potential inhibition of the complement activation in macular degeneration. First is the genetic picture. You just look from a historical perspective. I think around June in 2005 there were multiple publications in Journal of Science showing that polymorphisms in genes coding for complement regulatory proteins were involved in association with both the dry and the advanced – both the advanced dry macular degeneration and neovascular AMD as well. So when you look at that and you dovetail that with histopathologic studies both in primates and in humans, where you see that there’s significant staining for various components of the complement cascade, it’s reasonable to question if neovascular AMD, for example, or even dry AMD has complement mediated inflammation as one of its pathogenic phenomenon. And then when you look at systemic biomarkers of inflammation compared to age-matched controls, there were various studies showing that those inflammatory biomarkers are elevated. So taken together, it’s reasonable to postulate that complement activation has a role in wet AMD, and also dry AMD, of course, based on the genetic data. So we felt that anti-VEGF resistance comes in multiple flavors. One could be that the cell covering the endothelial cells in neovascular AMD, for example, imparts some anti-VEGF resistance, and another could be unabated inflammation, which may be complement mediate. So we did a study where we combined Zimura with Lucentis. At the time, I believe the Avastin data from Cap’s study was not available, and Eylea was not available commercially. And we did a study in roughly 40, 50 patients over a 6-month period, and really had compelling efficacy when compared to natural history. In the high does, for example, I think 60% of the patients gained three lines of vision. So based on the strength of the underlying science, the genetic association studies and the data that we had in albeit a smaller sample size with no control arm, we continued to develop that concept in the wet AMD. With respect to the dry AMD, as I stated, the genetic studies indicated that polymorphism confers a higher risk. When you couple that with the geographic atrophy as an endpoint that, at least this side of the Atlantic in the US, that expansion of geographic atrophy is an accepted endpoint. And we really don’t have any proven therapy for these unfortunate patients. So we started, I believe it was in 2010, we did approximately 48 patient study over a duration of a year just to test safety. And recall at that time the large amount of intravitreal injections that are administered for neovascular AMD. The safety profile and the acceptance wasn’t as great as it is today. So to do a one-year study in a patient who hasn’t had acute demise of vision as in geographic atrophy, we undertook, as I said, approximately 48 patient study over a year, and the safety was certainly very much acceptable. And we think that there was separation of curve which could indicate potential efficacy there, which obviously in lack of a controlled – there was no control arm and patients weren’t given the administration of the drug in a continuous fashion. But there was enough there that we’re progressing to a larger sample size study with a more controlled design.

TS: So are you confident still that the C5 is the right target for this approach?

SP: So if indeed complement mediated inflammation plays a role in both of those diseases, then the question is, you know, what’s the best place to – for antagonism of the cascade. And I don’t think anybody can definitively say what the right place to block is. But I think from a general perspective, most of the cellular damage that occurs from complement activation is because a membrane attack complex. And complement mediated inflammation now, there’s recent papers in neovascular AMD indicating that membrane intact complex really plays the major role. And so if that is the case, what you’d like to do is to block the formation of membrane attack complex and preserve some of the immune related functions in innate immunity that complement has in all of us. And that would at least theoretically give you the right safety and efficacy, safety to benefit ratio. And as far as inflammation is concerned, C5a is a mediator of inflammation. And by blocking C5, perhaps you’re not addressing C3a. But I think preclinical studies have shown that whether you block C3a or C5a, the emergence of neovascularization and its efficacy is roughly the same. So we believe that based on totality of the science, C5 is an ideal position to block if this pathway plays a major role in therapy for both diseases. But we’ll have to wait for the data to tell us what the right answer is.

TS: Sure. So there isn’t necessarily just one right answer in any of this. It could be C5a, could be C3a, could be multiple solution.

SP: We think membrane intact complex is the most important and the inflammation is also addressed by C5a. I guess one could make the point that C3 is a better target, or maybe some of the complement regulatory proteins upstream are better targets. But we feel confident that the ideal position, if this pathway is involved is downstream, and C5 is ideal if it’s downstream. But again, we’ll have to wait and see.

TS: Sure. And just finally, there were some eyebrows raised with Novartis, I think that they were going to continue LFG316 in dry AMD in combination with their anti-properdin. What do you believe is the rationale behind that statement, and do you have any sense of how successful it could be?

SP: I don’t think we can comment on that trial data or on that company. I mean I think when you evaluate data, you really need the totality of it to understand the statistical inferences that were made prior to the study, what the baseline variables were. And I just don’t have the totality of the data. But whenever you have an underpowered study in any situation, I think it’s very hard to draw conclusions one way or the other. But again, from our perspective, we just look at the data we’ve had and I think we do consider data from other situations. But we don’t see anything that tells us that you shouldn’t do larger powered studies with greater exposure over a longer period and assess in a controlled fashion.

TS: And last, we’re going to revisit our Podcast from last week. We spoke with Dr. Cedric Francois, the CEO of Apellis. And Apellis has built a pipeline of complement immunotherapies looking at several different indications, including AMD. In this interview, Dr. Francois explains why Apellis believes its approaches at better targeting C3 may have a better chance of success than those going after C5 or other parts of the complement system. Let’s have a listen.

Cedric Francois: The complement cascade is incredibly complex. I mean so you’re talking about something that probably rivals the clotting cascade in terms of complexity, where you have 3 pathways that can lead to activation. You know, the classical, the mannose binding lectin, and the alternative pathway. They all converge on complement factor C3. And then you get multiple effects of complement that come out of that. And the effects of complement can be split a little bit artificially to events that take place on the cell surface of target cells that are under quote-unquote attack by complement, versus what we call the fluid phase of complement, where factors get released in the interstitial fluid that play very important signaling roles in adaptive immunity. When in 2005, these snips were discovered in complement to be associated with macular degeneration, we had a whole bunch of stuff to figure out, right? One was, is this only associated with the incidence, or is it also associated with the progression of the disease? Is it associated with the early forms of the disease versus the late forms, you know, wet type versus geographic? All these things were unclear. And within that complexity, also the fact of which part of the complement affects the cell surface components, versus these fluid faced signaling components are really important? And to kind of put that in context, if we accept the premise that complement activation is bad for the retina in macular degeneration, then that can be bad either because complement attacks the retina directly, and that would be through the cell surface components. In that case, if you inhibit complement, you are really following a symptomatic approach, right, where you try to protect these retinal cells against insult by complement, versus on the other hand, complement playing an important role not as a direct aggressor of the retina, but a much more important role as a regulator of immunity in the back of the eye, where other immune elements, like macrophages, etc., come in and cause destruction through secondary mechanisms. The reason why these two are very different, and it’s not that they’re mutually exclusive, right, but they’re very different in terms of where you want to target the complement cascade if you want to affect these 2 sides of the equation. If you inhibit complement factor C5, you are very much making a bet on the direct insult to the retina, right, the attack on the cells, and the cell surface events? to be the problem in the disease. If you believe that immune regulation is the main contributor to the disease, then C5 is too far downstream in the cascade. You want to go upstream to complement factor C3, or for example, as Genentech does, complement factor D, because the alternative pathway accounts for approximately 80% of activation and products that contribute to the fluid phase. So when we approach macular degeneration with C3, we felt that we could cover all possible eventualities without knowing too much. I hate to admit it, but that’s kind of what happened. It’s not more scientific than that. But in the meantime, I think we found out much more of the past couple of years with a lot of evidence pointing in the direction for the role of the fluid phase and the immune regulation being important, and the role of complement playing a role there. Now we also start seeing now the first kind of, I’d say, circumstantial evidence from clinical trials that further seem to contribute to that hypothesis. With anti-C5 specifically targeted at those cell surface events, we actually had a study before Novartis came out that was run by Dr. Rosenfeld, who coincidentally is one of the co-organizers of this conference and this angiogenesis conference in Miami. He had already run a trial using Alexion’s Soliris, given intravenously to patients with geographic atrophy to find out if the progression of the disease could be slowed down. And that study was negative. So that was the first inkling that this C5 inhibition specifically might not be the right approach in geographic atrophy. But there was a sense that maybe intravenous was not the right route of administration, etc. And Novartis then ran the study looking at the intravitreal administration of their antibody against C5, which for all intents and purposes we can see is highly similar to Soliris. And now we know, unfortunately, you know, it was already read out about a year ago, I think, maybe even longer. But now we know that anti-C5 administered intravitreally also does not work. And now obviously that causes a lot of people to question, well, what about does that mean that complement in the retina is a losing proposition? And I think that is absolutely not the case. I’m obviously very subjective about this.

TS: Of course, you need to be. You’re the CEO of the company.

CF: Yeah. But also, looking as a scientist at this, and having probably done more deep work than a lot of other groups, there is a really important difference between where you target the complement pathways, vis-à-vis the effect that you can have. And you know, Genentech and Roche took a decision to move into phase 3 not purely based on hope, but I’d say on qualified optimism, right? And that qualified optimism stems from a deep understanding of how the fluid phase can affect adaptive immunity. And I will give you one simple example that points in that direction. When you look at macrophages in the retina, so in these diseases you can get monocytes that come into the tissue that sit there. You get macroglia that become involved. These cells will evoke stress on the retinal tissues and play a much more important role in contributing to the damage. Well, coincidentally, and again, this is an association, not causal, but it’s interesting, is that in wet type macular degeneration, the predominant phenotype of those macrophages are what we call M2 macrophages, which are kind of the reparative, pro-angiogenic, profibrotic macrophage phenotypes, which coincides with the clinical phenotype of wet AMD. Whereas in geographic atrophy, you see predominately M1 phenotype macrophages, which are the aggressive, damaging, all hell breaks loose type of microphage phenotypes, which coincides with, again, the clinical phenotype of GA. And there are a whole bunch of those associations that point in the direction of the role of the fluid phase. Which brings me to your last point, which is Novartis then decided not to abandon their anti-C5 program, but to instead combine their anti-C5 approach with an antibody against properdin. Now properdin is the only natural activator of complement that we have. So normally, complement gets activated by bacterial products and certain other elements, but nothing that comes out of our own body, typically, at least not in a natural context. Properdin is a complement factor that specifically drives complement activation. Now properdin is also something that’s interesting because it drives complement activation towards the membrane attack complex formation. So it is very interesting to see how this can play a role in the overall homeostasis of the retina. But one thing that I would like to point out is that 10 years ago – this might go a little bit too deep for this podcast, but yeah, I’m not going to too deep into that. I think suffice to say is that Novartis looked into doing molecular surgery within the complement pathways, where combining properdin inhibition, they believed, might make up for the lack of sufficient inhibition of complement by going too far downstream. And my only word of caution would be there that by inhibiting properdin, what you do is actually drive the complement activation away from the membrane attack complex, and the cell surface properties, towards the fluid phase. And it will be interesting to see if that truly makes the disease get better or not. To summarize, we’re far from knowing everything. I think between the read out of the phase 3 clinical trial by Genentech and our phase 2 clinical trial with the anti-C3 approach, we’re going to like get a lot of additional answers, with the final remaining question being is geographic atrophy too late for complement inhibition to have an effect, or do we want to go earlier in intermediate AMD, and is that where you want to block complement to avoid progression to the advanced form of the disease?

TS: Last question: how do you see this field playing out? Do you see a large percentage of the approaches being targeted today being successful? Or is there really going to be one moonshot that makes it all the way?

CF: I’d love to think that we are the moonshot.

TS: Of course you are. That’s what I was implying when I asked the question.

CF: Well, I think, look, I don’t believe in single moonshots. I think there’s always room for improvement. Beyond the mechanism, there’s also the product profile, right? I mean let’s say that complement inhibition is a good thing to do, and let’s say it’s a good thing to do in geographic atrophy. Then you don’t want to do an injection every month or even every other month, right? Then the question becomes is if you do complement inhibition, do you really need to do this chronically for many years, as is the case with anti-VEGF agents? Or can you really change the immune microenvironment where you can have long periods where you don’t need drugs anymore? And if that’s the case, where are we going to end up with good companion diagnostics, where we can actually identify patients or the state of the disease, I should say, based on blood biomarkers? So you can treat patients when it’s needed and not kind of – no pun intended, but blindly outside of that. So I don’t believe in a moonshot. I do think that with C3, again, subjectively, we have a good opportunity to look at broad complement inhibition. The difficulty that we face and everybody that works in this disease space is that we make these intravitreal injections, and we don’t really know whether we are truly inhibiting complement where it needs to be inhibited, right? We kind of run the experiment based on the maximum exposure that you can give. You inject as much as you can as often as you can, and you hope that it does what it was designed to do. But when studies are negative, is that because the drug doesn’t work? Hence, is it because the mechanism doesn’t work or is it because the drug didn’t get where it had to go, or wasn’t there for long enough? And that’s something on which, unfortunately, we don’t have enough of a handle.

TS: Well, there you have it, folks, our second KOL Corner. I hope you enjoyed this look at the complement cascade. Thank you to Dr. Jay Duker, Dr. Samir Patel, and Dr. Cedric Francois for joining us on this KOL Corner. Lots of excitement going on in the complement space. Every area has some bumps in the road. No doubt we look forward to this sector moving forward and bringing some relief, as Dr. Duker indicated, some relief for patients with macular degeneration who really, really need a lot of help. So thanks again for all of our KOLs and CEOs for joining us today. Thank you of course to our listeners for listening in. And don’t forget to go to ois.net. you can still register for the upcoming OIS@ASCRS. It’s happening on May 5. So go to ois.net, register, and we will see you in New Orleans.